Normal stem cells: The development of a robust in vivo model to study human beast epithelial morphogenesis by humanizing the mammary fat pad of NOD/ SCID mice as described by Kuperwasser et al. 2004 has been a major objective of our group. To humanize mammary glands with non-transforming cells, we used immortalized human mammary fibroblasts obtained from Dr. Kuperwasser. Additionally, we are developing our own immortalized, green florescent protein (GFP) tagged, human breast fibroblasts. Primary human fibroblasts from reduction mammoplasties are being successfully cultured in vitro. Cells have been infected with a retroviral construct encoding GFP and the catalytic subunit of human telomerase for immortalization. Once purified and selected these immortalized breast fibroblasts will be used for humanization of the mammary gland. Preceding injection into the humanized gland primary epithelial cells are cultured in a defined media reported to enrich for mammary stem cells. We have found that the addition of human prolactin to the defined media increases the number of putative mammary stem cells. Our preliminary data suggest that the cells also require a 3D structure, prior to injection, in order to successfully develop throughout the humanized gland. We found that co-culturing the epithelial cells with primary human mammary fibroblasts stimulates the formation of duct-like structures. Using glands collected in the initial trials, we have successfully developed the methods for distinguishing human fibroblasts from host murine fibroblasts. These methods include GFP whole-mount analysis as well as histological analysis using species-specific immunohistochemistry and genomic fluorescence in situ hybridization. Cancer stem cells: First we have attempted to repeat previously published data suggesting that a rare population of highly tumorigenic cells can be isolated from pleural effusions of women with stage 4 breast cancer. These putative cancer stem cells are unique in that they are the only tumor cells capable of initiating xenograft tumors in immunocompromised mice. We are also determining if heterogeneity in tumorigenicity exists within established breast cancer cell lines and identifying novel cancer stem cell markers. Work of Al Hajj et al. (2003) demonstrated that the cancer stem cells can be enriched from pleural effusions by the expression of the following antigens: Lineage- CD24-/dim CD44+. The Lineage cocktail consist of eight antibodies employed to ensure the removal of any non-tumor cells from the pleural fluid. Al Hajj observed that only Lineage- CD24-/dim CD44+ cells were capable of initiating tumors in mice, demonstrating their self-renewal properties. These authors also reported that the resulting xenograft tumors share a heterogeneous CD24 CD44 profile similar to that of the initial pleural effusion, demonstrating their multipotent nature. To date, we have injected cells from seven pleural effusions orthotopically into immunocompromised NOD/SCID mice. Cells from sample SH12 were sorted into two populations: Lineage-, CD24+ CD44+ and Lineage- CD24-/dim CD44+. As few as 200 CD24-/dim cells gave rise to tumors whereas 5,000 24+ cells failed to do so. The resulting xenograft had a CD24 CD44 profile similar to that of the initial pleural effusion with the majority of cells (> 80%) falling in the CD24+ CD44+ quadrant and the minority containing CD24- CD44+ cells consistent with Al Hajj. This xenograft has been successfully passaged in the mammary fat pad of mice multiple times. Resulting SH12 xenografts have been dissociated and sorted into four immunophenotypically unique Lineage- CD24 CD44 populations. These populations were injected into NOD/SCID mice to compare their tumorigenicity to that of the cell populations of the initial pleural effusion. These experiments are currently ongoing. Pleural effusion SH8 has been sorted and injected into mice in two separate experiments. In the initial experiment, three immunophenotypically unique Lineage- populations were injected (CD24bright CD44+; CD24-/dim CD44+; CD24+ CD44-). None of these populations resulted in tumors even when as many as 25,000 cells were injected. In a second experiment, SH8 was sorted into two populations, CD24bright CD44+ and CD24-/dim CD44+. As in the initial experiment, no tumors resulted even after injecting as many as 5,000 cells. These results are inconsistent with those reported by Al Hajj. Pleural effusion CC4 was sorted into three Lineage- populations (CD24bright CD44+; CD24-/dim CD44+; CD24+ CD44-) and injected into NOD/SCID mice. Similarly to SH8, no tumors formed with as many as 5,000 cells. These data suggest that the simple presence of Lineage- CD24-/dim CD44+ cells in a pleural effusion does not indicate the presence of cancer stem cells and that new markers for highly tumorigenic cancer stem cells are needed. We also have initiated an effort to generate data from primary breast tumors. Given the limitations of the generally small specimens received from lumpectomies, we are attempting to expand these in the mammary fat pad of immunocompromised mice. We have been successful with a tumor chunk received from a BRCA1 mutant carrier that has been successfully passaged in mice multiple times. We recently dissociated and sorted one xenograft tumor into two populations (CD24bright CD44+ and CD24-/dim CD44+) and injected them into NOD/SCID mice. This experiment is currently ongoing. We are currently determining whether there are cancer stem cells in established cells lines. First, we determined the CD44 CD24 profile of 23 breast cancer cell lines. We chose eight cell lines of varying in vivo tumorigenicity (MCF7, MCF10A-III, MCF10A-IV, ZR75-1, SUM229, SUM1315, SUM149 and AU565) for testing the effectiveness of using CD24 CD44 expression to enrich for highly tumorigenic cancer stem cells. To date, we have sorted MCF7, MCF10A-IV, ZR75-1, SUM229, and AU565 cells by CD24 CD44 expression and injected them into immunocompromised mice. No tumors formed in mice injected with the latter two cell lines, even when 10,000 CD24-/dim CD44+ cells were injected. Preliminary data from experiments injecting sorted MCF7, MCF10A-IV and ZR75-1 cells demonstrate that with the number of cells injected (1,000 to 10,000) all quadrants of the CD24 CD44 plot possess some tumorigenic potential. These early observations suggest that expression of CD24 (or lack thereof) has little influence on the ability of cells to form tumors. However, the trend is for cells lacking CD44 to be less tumorigenic than those expressing this protein. Additionally, we are working to identify novel markers for their ability to enrich highly tumorigenic cells. We are testing CD205, CD117, CD135, Cripto, SSEA4, CD87, CD200, CD133, CD29, CD49f, and CD177, chosen because they have been associated with tumor aggressiveness, embryonic stem cells, cancer stem cells in other tissues, and/or adult stem [summary truncated at 7800 characters]